If you’re familiar with RPGOT’s character optimization content, you’re familiar with my 4-tier rating system, but you might not know where it came from or why it works the way it does.
As a refresher, here’s the text as it appears at the top of all of my character optimization articles:
- : Bad, useless options, or options which are extremely situational. Nearly never useful.
- : OK options, or useful options that only apply in rare circumstances. Useful sometimes.
- : Good options. Useful often.
- : Fantastic options, often essential to the function of your character. Useful very frequently.
Where did this come from?
Short answer: I stole it. I’ve been wrongly credited with creating the 4-color rating system for years and I try really hard to clarify that this wasn’t my idea. I stole it from Treantmonk. (Don’t worry, he told me that he doesn’t mind)
I started RPGBOT some time in 2013, and at the time Treantmonk was basically the only big name in character optimization for Pathfinder. He had the most and the best character optimization content, and his rating system had mostly become a standard.
That’s not the only reason why I adopted it, but historical context helps.
Why 4 tiers?
I’ve seen a lot of other rating systems in character optimization content. 5 tiers is common, but I’ve also seen people use 7, 8, and 10 tiers.
I stuck to 4 tiers specifically because there is no “middle”. There’s no “this is fine” rating.
Character options aren’t like movies where if you see a bad one you’re out two hours and maybe the price of a ticket. If you pick something, you might be stuck with it for months or years depending on how long your game runs.
There’s little room for something to be “fine”.
When there’s a clear middle, ratings frequently form a sort of bell curve where the peak of the bell is in the middle and most things fall into that rating. If half of the things available are rated 3/5, it’s hard to tell what’s actually worth picking from those options.
As you stretch the rating scale to add more tiers, the distinction between each step gets smaller and less meaningful.
Ask yourself: if you’re considering a movie with an 8/10 rating and a movie with a 9/10 rating, how much better is the 9/10? Probably not very much. Now if you’re looking at a movie rated 3/5 compared to one rated 4/5, that’s still only one step on the scale, but it’s the difference between a movie that’s “fine” or “okay” versus one that’s “good”.
Why the colors?
Again: I stole the colors from Treantmonk. It’s a simple color pallet that’s easy for most people to distinguish at a glance, so people can easy scroll through articles and look for green and blue stuff.
But, as has been pointed out to me, if you have color vision differences, the color ratings can be totally unusable. For people who aren’t familiar with color vision differences, consider red-green color blindness: people with this condition have trouble distinguishing red and green from one another, and red and green are half of my color rating scale.
Why the stars AND the colors?
The star ratings are there primarily to make my content accessible for readers with color vision differences.
And don’t worry: the stars and colors always match. I’m using some code that I wrote to apply the colors and stars at the same time and I would have to break that code site-wide to make the colors and stars diverge.
How do you decide on ratings?
My short descriptions of the four rating tiers make it pretty clear what I’m looking for when I rate something.
I also try to do math when I can. Sometimes I can boil something down to numbers, and when I can do that it becomes much easier to justify a rating for something.
Will your rating system ever change?
No. I like it, and it has worked for 8+ years without change across several game systems. I suspect that it will continue to do so.
I have made very minor changes to the clarifying text over time to make the ratings more sensible for new readers, but the ideas behind the ratings have never changed and likely never will.
But I’m also not good at predicting the future. Assume that the answer is “no” until I say otherwise.